There are three types of strategies for maintaining data integrity:
This company needs to be certain that if a disaster were to occur it could recover very quickly. In order to do this, the recovery schedule it should use is the creation of a duplicate database. Because it has only one office, the duplicate database will have to be stored in another facility. Having both databases in the same facility would not protect the firm from the loss of data that could occur due to a fire or flood, which may destroy everything in one location. The firm may also develop backup power and communication resources and use RAID technology for fast recovery.
A before image is an image of a database record taken prior to the record being updated. An after image is an image of a database record taken after the record has been updated.
Because this corporation needs to operate its computer systems continuously (99.98 percent of the time) to continue to keep its good reputation, it needs a recovery scheme that involves duplication - mirroring with RAID. This would allow the corporation to recover in a matter of seconds if the database became corrupt or unusable by switching to the duplicate. You would expect to find full redundancy of the database in this case as well as redundant power, communication, and possibly computers.
Backward recovery is used to back out or undo unwanted changes to a database, whereas forward recovery is used to re-create a database using a prior database state. Both of these recovery methods require the use of journaling to capture the necessary data to undo or to re-create the database.
ROLLBACK is used to undo changes when something goes wrong in a transaction.
Locking at the row level allows other users to access other rows in the file while the locked row is being used. Locking at the page level locks all users out from being able to obtain access to any of the records that are included on the page of the record that is being updated. There is more resource conflict when locking at the page level than when locking at the record level. Locking at the record level also increases the processing overhead beyond what is necessary for locking at the page level.
There are three classes of authentication mechanisms. The first class is comprised of things a person knows or remembers, such as their name, account number, or password. The second class is things a person has or possesses, for example a badge or a key. The third class is personal characteristics that identify a person, such as their fingerprints or their signature.
Encryption is a transformation applied to data that makes it difficult to extract meaning. To implement it, you need an encryption technique to transform the data and a decryption technique to reconstruct the original data.
The advantage of using public key encryption is that it is easier to distribute the keys, because the public key can be distributed to anyone and only the holder of the private key can read it. Thus the private key does not need to be distributed eliminating the problem of distributing the sensitive keys.
Locking at the row level allows other users to access other rows in the file while the locked row is being used. Locking at the page level locks all users out from being able to obtain access to any of the records that are included on the page of the record that is being updated. There is more resource conflict when locking at the page level than when locking at the record level. Locking at the record level also increases the processing overhead beyond what is necessary for locking at the page level.
There are three classes of authentication mechanisms. The first class is comprised of things a person knows or remembers, such as their name, account number, or password. The second class is things a person has or possesses, for example a badge or a key. The third class is personal characteristics that identify a person, such as their fingerprints or their signature.
Encryption is a transformation applied to data that makes it difficult to extract meaning. To implement it, you need an encryption technique to transform the data and a decryption technique to reconstruct the original data.
The advantage of using public key encryption is that it is easier to distribute the keys, because the public key can be distributed to anyone and only the holder of the private key can read it. Thus the private key does not need to be distributed eliminating the problem of distributing the sensitive keys.
|
This page is part of the promotional and support material for Data Management (fifth edition) by Richard T. Watson |